Richard Arneson finds that Joel Feinberg's “principled arguments against hard paternalism are defective and merit rejection.” Furthermore, Arneson concludes with a tone of finality, “there is no successful case against hard paternalism to be made.” This diagnosis seems especially dire coming from Arneson, who formerly defended hard antipaternalism against Feinberg's revisionistic “soft” antipaternalism.¹

On Feinberg's behalf I will try to show how Arneson's critique downplays certain costs of hard paternalism. These costs may be worth paying, but (I think Feinberg believed) their magnitude itself constitutes a powerful case against paternalism.

In Arneson's view, Feinberg's antipaternalism is both an absolutism and a mystery. It is an absolutism because it denies that a certain consideration, voluntary harm to self, is ever a reason at all for criminalizing. It is a mystery because this very reason—voluntary harm to self—is not denied to be a reason for nonpenal regulation, nor for social correction. Penal sanctions, for Feinberg, constitute a distinctly graver kind of coercion; but it is still a mystery why the prospect of a sanction of a graver kind transforms what otherwise would be an admissible if slight reason into no reason at all.

Before addressing Arneson's criticisms, I think it will be worthwhile to unravel some of the mystery surrounding Feinberg's absolutism. Feinberg's liberal position limits criminalization to the prevention of wrongful harms and wrongful offenses—in short, to prevention of wrongs. The harm principle states that all but only wrongful harm to others is an invariable reason to criminalize.²

The offense principle states that all but only wrongful offense is an invariable reason to criminalize.³ The liberal position states that the harm principle and the offense principle exhaust the class of good reasons to criminalize. Ergo, Feinberg's liberal position entails that only what is wrongful may be criminalized.

Before considering Feinberg's mysterious formulation of the legal paternalist principle, let us note a direct antipaternalistic argument that his account of punishment and of the limits of criminalization would allow and that he on occasion seems to acknowledge. The master premises of this argument are these:

From these two premises and the liberal position, it follows that:

Of course, the liberal position, as a premise in the argument, needs support. And this is where the arguments against paternalism must come in. But Feinberg, in support, might simply argue that harm to self (if voluntary) is never wrongful even where there are reasons in favor of prohibiting it. Harm to self is never wrongful because harm to self (especially if voluntary) is not a wrong to self, nor—striking now against legal moralism—is it any other kind of wrong. This strategy would address a legal paternalism principle of this form:

Had Feinberg fixed this formula as his target, the antipaternalist's argumentative burden would be satisfied by merely showing that there are no wrongful voluntary self-harms, whether or not there are reasons to prohibit them. Feinberg could say this without surrendering to the tepid “balancing strategy,” legal paternalism, which holds that voluntary harm to self, though always a reason in favor of prohibition, happens never to add up to a sufficient reason.⁷

Feinberg could distinguish his view by saying that the volenti principle invariably blocks the passage from reasons against allowing voluntary self-harm to counting such harms as wrongs.⁸ Alternatively or additionally, Feinberg could deny on conceptual grounds that there is coherent sense to be made of self-wronging, as opposed to mere self-harming.

But, for intricate and indirect reasons,⁹

Feinberg's actual formulation of the legal paternalism principle leaves out the “wrongful” that is built into the harm and offense principles. “Harm” as it occurs in the legal paternalism principle means “setback to interests” rather than, as in the harm and offense principles, wrongful setback to interests. So the target is this:

So understood, the bare denial of legal paternalism is the assertion that a voluntary setback to one's own interests is sometimes not a reason in support of prohibition. But (for reasons still unclear to me) Feinberg gamely takes up as the antipaternalist's burden the seemingly impossible one of showing that a voluntary setback to one's own interests is never a reason in support of prohibition at all.¹⁰

There is at least one passage¹¹

in which Feinberg seems to have been aiming at legal paternalismW, which suggests that antipaternalism in Feinberg's spirit can mean either (or both) of the following:

The latter formulation, antipaternalism SI, gives effect to the idea that there is never a reason to prohibit an actor's voluntarily setting back the actor's own interest. Antipaternalism SI draws upon Feinberg's “want-regarding,” as opposed to “ideal-regarding,” concept of interests.¹²

Interests are closely connected to but not the same as wants;¹³ they are more stable and durable than mere desires, which, to become interests, must be realistic and capable of advancement by the agent's efforts.¹⁴ There are certain “general welfare interests” that for “almost all normal persons” promote their further interests, whatever those might be.¹⁵ Although there are interests an agent may not know he has—even final ends¹⁶—he may eliminate any of his interests by a psychological “withdrawal of investment”;¹⁷even general welfare interests may be extinguished. Reasons for action plausibly arise from interests and wither away with them.¹⁸

Arneson correctly points out that Feinberg's soft antipaternalism is a fight “on two fronts”¹⁹

against the hard paternalist and the hard antipaternalist, both of whom “agree in doubting that the idea of individual voluntary choice will bear the enormous weight” that Feinberg places on it.²⁰ Arneson's point is entirely on the mark, it seems to me. I think Feinberg's tendency to oscillate between antipaternalism SI and antipaternalism W is a sign that voluntary choice is buckling under the weight. The kind of wrongfulness-defeating voluntariness at work in antipaternalism W is ordinarily voluntariness of a relatively low grade and may have not much lifting to do should the concept of “self-wronging” turn out to lack sense. But the “withdrawal of investment” voluntariness at work in antipaternalism SI is both essential and of a high grade—especially where (as in the cases of suicide and self-maiming) the interest the actor wishes to cancel is a general welfare interest. The hard antipaternalist can insist that harm to self is never a wrong, whether or not voluntary, and that in any case, canceling the wrong of self-injury requires only a modest degree of voluntariness. The hard paternalist, on the other hand, can insist upon full voluntariness because canceling an interest of a high order puts the credibility of a purely “want-regarding” concept of interest under huge strain. Feinberg's “sliding scale” of “substantial voluntariness” has to shuttle frantically between these two fronts.²¹

But the instability of soft antipaternalism is not Arneson's main point; his point is, rather, the hopelessness of Feinberg's arguments against hard paternalism, which rest upon an appeal to an idea of personal sovereignty. Arneson makes a series of telling points that, I think, manage to uncover the power of Feinberg's underlying conception even as they expose its limitations.

Arneson says, “The personal sovereignty idea,” properly understood, “postulates a moral right; it does not say there can be no valid moral reasons that conflict with the right.” All rights are subject to infringement short of violation where the interest protected by the right is set back in small degree and the interests advanced by the infringement are great: “What is counterposed to the right against paternalistic interference is always the good of the individual whose right we are considering infringing,” Arneson says. “But in a particular case, the good of the individual that is at stake can be enormous and…the agent's interest in self-determination…very slight…. Seen in this light, absolutist antipaternalism…is fanaticism,” he concludes.²²

But—contra Arneson—it is precisely this idea of a “good of the individual” distinct from what she competently posits as her good that Feinberg's “want-regarding” concept of interests opposes. The right tracks the interest. And fanaticism, from Feinberg's perspective, consists in imputing to an agent a “good for her” that she competently rejects. Arneson expresses the target of Feinberg's concern precisely when he says, “The essence of paternalism is overriding the individual's own evaluation of where her own good lies.”²³

Who's life is it, anyway? Arneson says it is an “exaggeration” to claim that “a person's fully voluntary choices uniquely reflect the agent's settled values, attitudes, and desires”²⁴—but to say this is to lend credence to an “ideal-regarding” concept of interests, insofar as one's interests are no longer synonymous with one's settled values, attitudes, and desires.

According to Arneson: “A person may on some occasion voluntarily choose to act against settled values, attitudes, and desires,” which themselves may include wanting occasionally to be a “wild and crazy,” less-than-fully-voluntary guy who acts against character.²⁵

Feinberg need not deny these possibilities; nor does admitting them alter the implication of Feinberg's presentistic, “want-regarding” concept of interest. If what an agent volunteers to do now is to be a man of courage who insults his dean, that is what he becomes—it is not in his interest to be restrained even though he will regret it tomorrow. It is for the agent to determine the import of his precommitments and the weight to be given to his settled habits and aspirations and possible future maturation and transformation. If the agent structures these so that the “wild and crazy” guy is lashed to the figurative mast, it is in his interest to be restrained. But not otherwise. If this position is an exaggeration, it is only because a “want-regarding” concept of interests is an exaggeration. And to depart from presentism is to idealize interests in a way that will be hard to contain. Any departure worth making will have reference not only to what the actor would eventually want but to what she would want were she fully informed, mature, free of certain cognitive biases, and so on. And this leads us far down the path toward simply stipulating what she should want. Do we wish to adopt an “ideal-regarding” concept of interest to avoid the consequences of “want-regardingness”? I say we do, but with awareness—thanks to Feinberg—of the full costs of doing so.

Arneson acknowledges that Feinberg's “subjectivism” about the good makes antipaternalism “somewhat more plausible”²⁶

but cautions that “even taking into account the crucial value of autonomy, it remains the case that sometimes a hard coercive shove away from the bad can improve anyone's life.”²⁷ But Feinberg need not deny that a violation, small or large, of personal sovereignty may bring enormous personal benefits—and be so acknowledged by the agent. What Feinberg denies is that the benefit justifies the violation.²⁸ What benefits me at some later time need not have been in my interests—as I had determined them—at the time they were forced upon me. “You'll thank me later,” on Feinberg's view, is never an excuse unless what the coercer is to be thanked for is for having prevented a failure of voluntariness at the relevant moment.

According to Arneson, Feinberg “fails to notice a crucial truth that is in the neighborhood of the Kantian assertions. This truth…sink[s] Feinberg's ideal of personal sovereignty.”²⁹

This truth is a “secular sanctity-of-life” doctrine that “incorporates the idea that rational-agency capacity has special value, for it is this that makes the human person's life normally a great opportunity.”³⁰ Nonetheless, the doctrine acknowledges that as life dwindles, its value dwindles also. This truth enables us correctly to prohibit a Pouting Young Adult from suicide while allowing suicide at life's end to alleviate pain. Kant would prohibit both, while Feinberg would forbid prohibiting either.³¹ Although, in Arneson's view, Feinberg wrongly disparages such a doctrine, it is intelligible as a “significant obligation” akin to trusteeship. As Arneson puts it, “My life is mine to live, but not in any way I might want or arbitrarily choose. I owe a duty to the moral community to use well the opportunity…not to waste it or let it spoil.”³²

Feinberg would applaud this doctrine as a personal ideal, but he would also point out that if Arneson's secular value is supposed to qualify every agent's interests, willy-nilly, it exemplifies an “ideal-regarding” concept of interest and in effect defines a sort of “moral harm” from which the agent may be protected.³³

Enforcing it via the criminal law would be rejected by Feinberg as a case of legal moralism rather than legal paternalism. Insofar as the duty Arneson contemplates is not merely self-regarding but also one owed to the moral community as such, Feinberg no doubt would warn that if I hold my life in an unconsented trusteeship, subject to the moral community's insistence that it not be wasted or squandered, it is hardly “my life to live” anymore. Moreover, if the objective value of my life diminishes over time in a way that relaxes my duty not to waste it, it is not immediately obvious why the duties of others with respect to me (both of aiding and of not harming) would not dwindle coordinately—and similarly, if circumstances prematurely curtail my capacity to contribute to society or to make something of my life.

Managing or avoiding such difficulties will be among the tasks of the “watered-down secular equivalent” of a sanctity-of-life doctrine that Arneson outlines. The undertaking is worthwhile because Arneson's responses to the two cases—the Pouting Young Adult and the terminally ill sufferer—are intuitively correct. It is a huge and difficult undertaking, but such are the costs of any principled defense of paternalism. We are fortunate to have had Joel Feinberg to show us how to go about any project of formulating, probing, defending, and reformulating the principles that shape our abiding moral responses.